Monster Mash
Last updated: 2021-09-09
Checks: 6 1
Knit directory: natarajanlab_wiki/
This reproducible R Markdown analysis was created with workflowr (version 1.6.2). The Checks tab describes the reproducibility checks that were applied when the results were created. The Past versions tab lists the development history.
The R Markdown is untracked by Git. To know which version of the R Markdown file created these results, you’ll want to first commit it to the Git repo. If you’re still working on the analysis, you can ignore this warning. When you’re finished, you can run wflow_publish to commit the R Markdown file and build the HTML.
Great job! The global environment was empty. Objects defined in the global environment can affect the analysis in your R Markdown file in unknown ways. For reproduciblity it’s best to always run the code in an empty environment.
The command set.seed(20210512) was run prior to running the code in the R Markdown file. Setting a seed ensures that any results that rely on randomness, e.g. subsampling or permutations, are reproducible.
Great job! Recording the operating system, R version, and package versions is critical for reproducibility.
Nice! There were no cached chunks for this analysis, so you can be confident that you successfully produced the results during this run.
Great job! Using relative paths to the files within your workflowr project makes it easier to run your code on other machines.
Great! You are using Git for version control. Tracking code development and connecting the code version to the results is critical for reproducibility.
The results in this page were generated with repository version d78e0cd. See the Past versions tab to see a history of the changes made to the R Markdown and HTML files.
Note that you need to be careful to ensure that all relevant files for the analysis have been committed to Git prior to generating the results (you can use wflow_publish or wflow_git_commit). workflowr only checks the R Markdown file, but you know if there are other scripts or data files that it depends on. Below is the status of the Git repository when the results were generated:
Ignored files:
Ignored: .DS_Store
Ignored: .Rhistory
Ignored: analysis/.Rhistory
Ignored: analysis/metaplot_strat_torus.nb.html
Ignored: analysis/monster_mash.nb.html
Ignored: analysis/testplot.nb.html
Untracked files:
Untracked: analysis/metatrials_sim.Rmd
Untracked: analysis/monster_mash.Rmd
Untracked: analysis/testplot.Rmd
Untracked: analysis/torus.Rmd
Unstaged changes:
Modified: analysis/metaplot_strat_torus.Rmd
Note that any generated files, e.g. HTML, png, CSS, etc., are not included in this status report because it is ok for generated content to have uncommitted changes.
There are no past versions. Publish this analysis with wflow_publish() to start tracking its development.
f=list()
files <- list.files(pattern = "\\.tsv")
for(i in 1:length(files)){
a=fread(files[i],header=T,sep="\t")
pheno=strsplit(files[i],split="-")[[1]][2]
id=paste0(a$chr,":",a$pos,":",a$ref,":",a$alt)
df=data.frame(id,a$beta_meta,a$se_meta)
rm(a)
colnames(df)=c(paste0("id_",pheno),paste0("beta_",pheno),paste0("se_",pheno))
f[[i]]=df
}
##venous embolism is not in anything but EUR
i=5
a=fread(files[i],header=T,sep="\t")
pheno=strsplit(files[i],split="-")[[1]][2]
id=paste0(a$chr,":",a$pos,":",a$ref,":",a$alt)
df=data.frame(id,a$beta_EUR,a$se_EUR)
rm(a)
colnames(df)=c(paste0("id_",pheno),paste0("beta_",pheno),paste0("se_",pheno))
f[[i]]=df
saveRDS(f,”pheno_files.rds”)
###switch to lipids
files <- list.files(pattern = "\\.tsv")
files
f=list()
for(i in 1:length(files)){
a=fread(files[i],header=T,sep="\t")
pheno=strsplit(files[i],split="-")[[1]][2]
print(pheno)
id=paste0(a$chr,":",a$pos,":",a$ref,":",a$alt)
df=data.frame(id,a$beta_meta,a$se_meta)
rm(a)
colnames(df)=c(paste0("id_",pheno),paste0("beta_",pheno),paste0("se_",pheno))
f[[i]]=df
}
allfiles=readRDS("master_pheno.rds")
zmat=data.frame(matrix(, nrow=nrow(allfiles), ncol=(ncol(allfiles)-1)/2))
for(i in c(2,4,6,8,10,12,14,16)){
z=allfiles[,i]/allfiles[,i+1]
zmat[,(i/2)]=z
}
250.2: phecodeType 2 diabetes
401: phecode Hypertension
411.4:phecode Coronary atherosclerosis
443 peripheral vascular
452 Other venous embolism and thrombosis
continuous-LDLC-both_sexes-medadj_irnt: LDLC
biomarkers-30870-both_sexes-irnt (Triglycerides)
biomarkers-30760-both_sexes-irnt (HDL)znew=readRDS("~/Downloads/lipidffiles/z_allpheno_naomit.rds")
bed=read.table("~/Dropbox//ld_chunk.bed")
head(bed)
maxes=apply(znew[,c(5:12)],1,function(x){max(abs(x))})
znew$maxes=maxes
bed$V1=do.call(rbind, strsplit(bed$V1, 'chr'))[,2]
max_block=data.frame(matrix(ncol = ncol(znew)+1, nrow = nrow(bed)))
colnames(max_block)=c(colnames(znew),"snpid")
for(i in 1:nrow(bed)){
chr=bed[i,1]
start=bed[i,2]
stop=bed[i,3]
in_chrom=znew[znew$a.chr==chr,]##extract those of the same
goodguys=in_chrom[in_chrom$a.pos>start&in_chrom$a.pos<stop,]
if(nrow(goodguys)>0) {
z.max=which.max(goodguys[,"maxes"])
z_good=goodguys[z.max,]
} else {
z_good=rep(0,ncol(max_block))
}
z_good=data.table(z_good,stringsAsFactors = T)
z_good$a.chr=as.character(z_good$a.chr)
#z_good$=rownames(goodguys)[z.max]
max_block[i,]=c(z_good,rownames(goodguys)[z.max])
print(i)
}
max_block=na.omit(max_block)
saveRDS(max_block,"monster_mash_max_block.rds")source('~/Dropbox/mash_materials/flashscript.R')
#zmash=readRDS("~/Dropbox/cadproject/zmash_mat.rds")
# identify a random subset of 20000 tests
random.subset = sample(1:nrow(zmash),40000)
zmash=as.matrix(znew[,c(5:12)])
data.temp = mash_set_data(zmash[random.subset,],alpha = 1)
Vhat = estimate_null_correlation_simple(data.temp)##correlation matrix of the errors
saveRDS(Vhat,"~/Dropbox/monster_mash/Vhat.rds")
library("lattice")
clrs = colorRampPalette((c("#D73027","#FC8D59","#FEE090","#FFFFBF", "#E0F3F8","#91BFDB","#4575B4")))(64)
print(levelplot(Vhat,col.regions = clrs,xlab = "",ylab = "",colorkey = TRUE,main="VHAT"))
rm(data.temp)
data.random = mash_set_data(zmash[random.subset,],alpha = 1,V=Vhat)##use for E(Z)model
max_block=readRDS("~/Dropbox/monster_mash/monster_mash_max_block.rds")
zmax=apply(max_block[,c(5:12)],2,function(x){as.numeric(x)});
rownames(zmax)=max_block$snpid
data.strong = mash_set_data(zmax,alpha = 1,V=Vhat)
U.pca = cov_pca(data.strong,3)
U.flash=cov_flash(data.strong, non_canonical = TRUE)
X.center = apply(data.strong$Bhat, 2, function(x) x - mean(x))
U.ed = cov_ed(data.strong, c(U.pca,U.flash,list("XX" = t(X.center) %*% X.center / nrow(X.center))))
saveRDS(U.ed,"~/Dropbox/monster_mash/EDcov.Rds")
U.c = cov_canonical(data.random)
m = mash(data.random, Ulist = c(U.ed,U.c),outputlevel = 1)
saveRDS(m,"~/Dropbox/monster_mash//mfitMONSTER.rds")znew=readRDS("~/Dropbox/monster_mash/z_allpheno_naomit.rds")
zmash=as.matrix(znew[,c(5:12)])
m=readRDS("~/Dropbox/monster_mash/mfitMONSTER.rds")
k=length(m$fitted_g$Ulist)
l=length(m$fitted_g$grid)
pimat=matrix(m$fitted_g$pi[-1],nrow=l,byrow=T)
colnames(pimat)=names(m$fitted_g$pi)[2:(k+1)]
barplot(colSums(pimat),las=2)
library("lattice")
for(i in 1:7){
z.num=as.matrix(cov2cor(m$fitted_g$Ulist[[i]]))
colnames(z.num)=colnames(zmash)[c(5:12)]
rownames(z.num)=NULL
clrs = colorRampPalette(rev(c("#D73027","#FC8D59","#FEE090","#FFFFBF", "#E0F3F8","#91BFDB","#4575B4")))(64)
z.num[lower.tri(z.num)] = NA
print(levelplot(z.num,col.regions = clrs,xlab = "",ylab = "",colorkey = TRUE,main=paste0(names(m$fitted_g$Ulist)[[i]])))
}
head(z.num) beta_452 beta_30760 beta_30870 beta_LDLC <NA> <NA> <NA>
[1,] 1 0.466683 0.4664710 0.5046696 0.21310930 -0.49137031 0.5169558
[2,] NA 1.000000 0.6501601 0.7198942 0.01171034 -0.27029396 0.2059288
[3,] NA NA 1.0000000 0.9481645 0.04822896 -0.29407272 0.2840662
[4,] NA NA NA 1.0000000 0.05447762 -0.24944494 0.2104635
[5,] NA NA NA NA 1.00000000 -0.07529556 0.0834628
[6,] NA NA NA NA NA 1.00000000 -0.7830241
<NA>
[1,] 0.13933247
[2,] -0.12810318
[3,] 0.07842253
[4,] -0.01957032
[5,] 0.02615585
[6,] -0.29857400Now after computing posteriors (on the cluster), we examine results
#zmash=readRDS("~/Dropbox/monster_mash/z_allpheno_naomit.rds")
pm=readRDS("~/Dropbox/monster_mash/posterior_means.rds")
lf=readRDS("~/Dropbox/monster_mash/lfsr.rds")
ptab=2*pnorm(-abs(zmash))
sum(lf<0.05)[1] 2974480sum(ptab<5e-8)[1] 274177r=sample(nrow(pm),100000)
plot(zmash[r,],pm[r,],xlab="MLE",ylab="MASH",main="MashEstimatesvsZ")
sum(lf<0.05)/sum(ptab<5e-8)[1] 10.84876lf